DESCRIPTION: Protein engineering is a powerful tool for design if novel liquid crystal phases, macromolecular surface arrays, reversible hydrogels, and artificial extracellular matrices for use in tissue regeneration and repair. In vivo microbial expression of artificial genes provides a means of preparing such non-natural proteins in high yields. The target structure is encoded into an artificial gene, and the gene is expressed in an appropriate microbial host. However, in vivo protein engineering poses a challenge in that the pool of potential monomers is restricted to the natural proteinogenic amino acids and those analogs that can be activated and charged to transfer RNAs. Tirrell and others have successfully incorporated analogs of methionine, isoleucine, leucine, and phenylalanine through the action of their respective aminoacyl-tRNA synthases. Analogs with olefinic and acetylenic functional groups have been shown to serve as methionine surrogates in bacterial protein synthesis. Incorporation of such functional groups creates important new opportunities for chemical derivatization, extending the range of materials properties that can be designed into protein-based polymers. For example, recent advances in the chemistry of olefin metathesis have led to the development of transition metal carbenes that catalyze efficient cyclization of peptides containing olefinic side chains. The objective of this proposal is to combine fast computational analog screening methods and experiments - both in vivo and in vitro - to find new amino acids for use in protein engineering. This collaboration will lead to fast and efficient discovery of non-natural amino acid analogs with new and useful functionality and will provide a basis for building novel protein-like polymers with desired properties. The computational methods to be used here have already been tested for design of analogs for phenylalanine and will be extended to new substrates for Phe, Met, Ile, Leu, and Val tRNA synthetases.